I decided to investigate whether the weekly high and low was more likely to occur on a Monday and Friday for other markets the same as the for the XJO. Looking at the HSI it would seem to hold the same pattern as the Australian XJO index. Using data from the start of 2008 it would seem to hold the same probability of printing the high and lows at either end of the week, roughly twice as likely then Tuesday, Wednesday or Thursday.
Clearly Mondays without other information or biases is hard to guess whether the high or low is the one you should be fading but once the weekly trend is set it doesn't make sense to fight the trend within the week. It seems weekly momentum has a bias to carrying on to the end of the week.